Talk:Two dots (diacritic)/Archive 1

Archive 1

Who borrowed what from whom?

The historical introduction says:

[...] in the forms of handwriting which emerged in the early modern period (of which Sütterlin is the latest and best known example), the letter <e> had two strong vertical lines, and the superscript <e> looked like two tiny strokes. Gradually these strokes were reduced to dots, and as early as the 16th century we find this handwritten convention being transferred sporadically to printed texts too.

But the section on the diaeresis mark has the following:

In Greek, French, Dutch and Afrikaans, the diaeresis is expressed by putting the diaresis mark or trema over the second of the two adjacent vowels, as in the (Greek) names Chloë and Zoë. The term trema, as the diacritical symbol itself, is taken from Medieval Greek Byzantine orthography, trema, meaning "perforation, orifice".

Byzantine Greek is much older than the 16th century.

Given this, I don't think it can be claimed, for example, that Spanish borrowed the trema from the German umlaut. A borrowing from medieval Greek seems much more likely. I am editing the article to be more neutral on this matter.

The claim that Eastern European languages written with the Latin alphabet such as Estonian borrowed the umlaut sign from German is more plausible, but it still needs to be checked. FilipeS 21:12, 15 May 2006 (UTC)

Quite correct. Umlaut was developed in German in Early Modern Times as an entirely new indigenous phenomenon. It happens to look like the far older trema/diaeresis. I think that some Scandinavian and Eastern European languages may have borrowed Umlaut from German, and this is certainly true of Turkish, which switched from the Arabic alphabet to a carefully adapted variety of the Latin alphabet in the early 20th century. But the trema found in French and other languages which indicates the separation of vowels rather than a vowel fronting is entirely unrelated. An earlier version of this article distinguished these clearly, but I see someone has edited in such a way that it looks like umlaut, trema and diaeresis are synonyms. Do change this, and please remove all information about trema from this article, except for a brief cross-reference to the trema article. --Doric Loon 05:36, 16 May 2006 (UTC)

But the diacritics are one and the same. They're recorded the exact same way in the computer, they're represented by the exact same piece of hot lead, and I seriously doubt that just about any printer ever worried about the difference. I actually doubt that they were an entirely new phenomenon; 18th century and 19th century German printers typeset Greek on a regular basis, for one thing. At the very least, they've been conflated in modern times into one diacritic, both represented by U+0308 in Unicode. Insisting that they are completely different simply doesn't reflect reality.--Prosfilaes 06:24, 16 May 2006 (UTC)

Well, no, they are slightly different in form. The article says:

In professional typography, the umlaut dots are usually a bit closer to the letter's body than the diaeresis dots, while in handwriting, and in most computer screen fonts, no difference is made.

and:

The need to distinguish between the umlaut sign and the trema in Unicode has led to the following recommendation by ISO/IEC JTC 1/SC 2/WG 2...

So they are different in modern typographical usage, though I don't suppose most readers would care much about it. But what is more important is the history, since this article is also partly devoted to that, and this requires two separate discussions. Because on this one point you are wrong, Prosfilaes: the origins of the Umlaut diacritic are very precisely known. Above all, though, the two diacritics serve completley different linguistic functions. It is not helpful to muddle the discussions. Why not make something of the article diaeresis (diacritic), which at present is a redirect? --Doric Loon 09:57, 16 May 2006 (UTC)

I've added a citation request to the "In professional typography"; one message on the Unicode mailing list implies that German typesetters did indeed make the distinction, but that's not citable. As for the need to distinguish between the two, that came from the German library network, which shows just what your arguments already have; that there's some who believe they need seperating. --Prosfilaes 19:40, 16 May 2006 (UTC)

Doric Loon, you can propose a split, but I would disagree with such a proposal, for the following reasons:

• While the umlaut mark and the trema are differentiated in some typographical conventions, they still look very similar: two dots over a letter.
• Not everyone distinguishes the umlaut mark from the trema. Computer software usually does not differentiate the two.
• The umlaut mark and the diaeresis mark do not represent just two opposed phenomena, umlaut and diaeresis. In many languages now, the double dots indicate a change of pronunciation which is neither an umlaut, nor a diaeresis, in the grammatical and phonological sense. Historically, these symbols may have started out as indicators of umlaut/diaeresis, but now their value has been broadened. Why not broaden it a little further, and treat the two in the same article? (Referring, of course, that some typographical conventions distinguish the German umlaut from other types of double-dots.)
• You wrote that the umlaut is an entirely indigenous German development. I don't doubt that early umlaut signs were. But those superscrip es over vowels were not quite like the modern umlaut sign. And, for many centuries, intellectuals all across Europe knew French, which uses the trema... Could there not have been an influence of French on German, which favored the evolution towards the two dots? FilipeS 14:38, 16 May 2006 (UTC)

Your last point is an interesting one, which I don't suppose anyone can disprove. OK, I am not all out for a page split, but let's make sure that within the page we differentiate the traditions clearly. And we could consider calling the page "Umlaut and diaeresis (Diacritics)". --Doric Loon 16:10, 16 May 2006 (UTC)

I also disagree with splitting the article, because of the reasons FilipeS mentioned, but I do like your move proposal. I almost did the move myself. But your consideration might be controversion, so I follow Requested move procedure. — Adhemar 18:43, 8 August 2006 (UTC)

BTW, FilipeS, I've just seen your rewritten intro. That seems a big improvement. And generally, thanks for the massive amount of work you've done here recently. --Doric Loon 16:22, 16 May 2006 (UTC)

Alphabetical order

In Swedish, ä and ö (and å) come at the end of the alphabet, after z. - - - 21 July 2006

The following discussion is an archived debate of the proposal. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the debate was No consensus. —Wknight94 (talk) 01:27, 27 August 2006 (UTC)

Requested move

• Umlaut (diacritic) → Umlaut and diaeresis (diacritics) … Rationale: The article is both about the umlaut mark and the diaeresis mark. A split has been suggested, but many oppose. However, the title should reflect both uses. Umlaut (diacritic) and Diaeresis (diacritic) should redirect to this. Both Umlaut and Diaeresis should disambiguate between the diacritic and the linguistic usage of the terms. — Adhemar 18:41, 8 August 2006 (UTC)

Survey

• Support obviously, as I started this procedure myself — Adhemar 18:48, 8 August 2006 (UTC)
• Oppose: Putting the names together into one title using "and" is not standard naming. If the article ought to be forked, then so be it; apparently someone merged it in the past [1]. If not, the title is not the place to put various names of similar things. If this article is to have both, then Diaeresis belongs as a redirect here to Umlaut (diacritic). —Centrx→talk • 01:11, 27 August 2006 (UTC)

Discussion

See consideration of Doric Loon

The above discussion is preserved as an archive of the debate. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

L'Haÿ-les-Roses

This article states that the in "ÿ" in L'Haÿ-les-Roses is a modified form of "ï", whereas the Heavy metal umlaut page says it is a modified form of "IJ". Then on the actual L'Haÿ-les-Roses page, it says that the city was named after Roseraie de L'Haÿ, and that article doesn't mention "L'Haÿ" being an alternate spelling/typography for any other word or letter. Can someone clear this up? --Psiphiorg 04:45, 27 September 2006 (UTC)

I've modified the Heavy metal umlaut page. The ÿ is not a modified form of ij, the two just look very similar (and can be indistinguishable in handwriting). See IJ (letter) for the history of ij. As to stating that ÿ in is a modified form of ï, I find that very sloppy/unclear. It can be true though, depending on what you mean by modified form. — Adhemar 08:34, 27 September 2006 (UTC)

Capitalized Umlauts on the Swiss keyboard

I would edit directly, but me written english may be inaccurate.

The Swiss keyboard has the same visible layout for swiss-french and swiss-german.

That means pushing the key "ö/é" with keyboard-driver set on swiss-german leads to the umlaut "ö". Which a keyboard-driver set to swiss-french, the same key leads to "é". Pushing the key in combination with "Shift", the effect is vice versa.

The add a capitalized umlaut, one may use the procedure written in the current version of the article. But lots of us are using the key-sequence "Caps Lock On" - A - "Caps Lock Off", which leads to "Ä".

Jackobli 19:03, 5 December 2006 (UTC)

Previous talk

Should have added this earlier. There are older discussions at Talk:Diaeresis, from before the split. FilipeS 17:59, 6 February 2007 (UTC)

Anais

The article states that Anais would be pronounced [a'nɛs]. I disagree, I'm convinced it would have to be [a'nɛ], as in Marais. --BjKa 08:44, 16 February 2007 (UTC)

Since the word does not actually exist, there is no true answer. However, the point of the example is to compare the pronunciation of the pair of letters ai in Anaïs and Anais. Removing the [s] from the pronunciation would likely just confuse the readers. FilipeS 13:15, 16 February 2007 (UTC)

Why not like in Calais, pronounced [ca'lɛs]?--Prosfilaes 13:29, 16 February 2007 (UTC)

Isn't "Calais" pronounced [kaˈlɛ]? FilipeS 19:27, 22 February 2007 (UTC)

Looking at the context of this statement, in French it usually would be. Since Calais is an English city cruelly held by the French in violation of international law, however, its proper pronunciation is [ca'lɛs].--Prosfilaes 19:32, 22 February 2007 (UTC)

ï

Yes, the ï, I mean, I've seen it used in Saïx's name, but not sure on how it works out properly. I mean, does it go to repeart the a and ï in pronounciation or what? Cause when I pronounce it right, it sounds like it's doing the "ai" twice, by saying "Sai-ax" sort of way. Is that correct on the matter? Cause I had a bit or a read through the article, but couldn't really find coverage on that part. Captain Drake Van Hellsing 04:45, 21 April 2007 (UTC)

Trema

Is there a particular rationale why "trema" is listed first in the article? The English name of the diacritic is "diaeresis", not "trema". "Tréma" is the French name for the diacritic which English calls "diaeresis".—Gniw (Wing) 17:46, 5 February 2007 (UTC)

I don't believe it's that simple. [2] uses trema as an English word; this post on the Linguist List does also, [3] says "The “diaeresis” (Coptologists generally prefer the term “trema” in view of the various functions of thischaracter in Coptic in comparison with Greek". A forum poster says "The symbol itself (the two dots above) is called a trema. ... "Trema" is a term for the symbol used as diæresis and umlaut, used by typographers and pretty much nobody else.". "Trema" certainly is an English word, used as part of a triple collection of words, "diaeresis", "trema" and "umlaut". When all three are distinguished, "umlaut" is consistenly used for only some usages of the two dots, but which one of "trema" and "diaeresis" is used to name the symbol and which one names the non-umlaut usages seems to depend on the author.--Prosfilaes 20:39, 5 February 2007 (UTC)

I wouldn't place too much trust on what the Unicode Consortium says, since for some things (such as the Big5 mess, and partly for this umlaut–diaeresis mess, to mention just two examples) it is quite a large part of the problem itself. (Given its perceived credibility, even authoritativeness, it has the power to do a lot of damage very easily; the same goes for Wikipedia, IMHO).—Gniw (Wing) 20:55, 5 February 2007 (UTC)

Words have meaning based on how they are used. It's clear that people are using trema as an English word, and only one of the cites above was to Unicode.--Prosfilaes 21:38, 5 February 2007 (UTC)

Diaeresis is definitely the established term in English. "Trema" does not even have an entry in the Concise Oxford Dictionary, nor in dictionary.com (except with a totally different and irrelevant meaning), nor on m-w.com (unless you search the unabridged, and even then I don't know whether it's used with the French meaning). The established English term "diaeresis" should therefore be given precedence. -86.140.131.100

It is given precedence. That's what the article uses most of the time. FilipeS 21:04, 29 April 2007 (UTC)

OK. I was referring to the 28th April version of the article, which began "The umlaut mark (or simply umlaut) and the trema or diaeresis mark (or simply diaeresis)" and (in the following paragraph) "The trema or diaeresis". I agree with the changes you've made since that time. Thanks. --86.140.131.100 21:27, 30 April 2007 (UTC)

Guarani

Do you know that Guarani uses the diacritic in the letter g? It would be interesting that you put it in your article.

As far as I know standard Guarani uses the tilde over the letter "g", not the umlaut. FilipeS 01:19, 22 June 2007 (UTC)

Merging with Umlaut

I think this article should be merged with Umlaut. What do you think? Peteturtle 10:09, 17 August 2007 (UTC)

(edit) Never mind. I didn't notice that was a disambiguation page. Peteturtle 10:09, 17 August 2007 (UTC)

Rename article? / own lemma for diaeresis

At the moment, this article is named Umlaut (diacritic) and both Trema (diacritic) and Diaeresis (diacritic) redirect here. I think Umlaut should not be the primary name for this article: it should be called Diaeresis and Umlaut can link to it. This is also more in line with the Diacritical marks template.

My primary reason for changing its name would be that Umlaut refers to the use of this diacritical mark in German and Diaeresis is broader.

Richard 08:58, 17 September 2007 (UTC)

That would be no improvement - the article is about two phenomena, which we take together because they look similar, and calling the article by the name of one of these is therefore a little odd, but switching to the other is no better. We discussed this at length, and you should read the above before you move forward, but I still think "Umlaut and diaeresis" would be the best title. Doric Loon 13:33, 17 September 2007 (UTC)

Unfortunately, Wikipedia's style guidelines advise against double titles with "and". Incidentally, Diaeresis is already taken. It would have to be Diaeresis (diacritic). Perhaps in the future the article can be split into Umlaut (diacritic) and Diaeresis (diacritic). It's getting a little large. FilipeS 20:40, 17 September 2007 (UTC)

Well, I supported splitting last time, but was soundly outvoted. Is that wiki guideline really inflexible? Sometimes I would have thought a lemma with "and" in the middle was exactly right. --Doric Loon 21:42, 17 September 2007 (UTC)

I wouldn't go so far as to split the article. My proposal was just to swap the primary name with that of a redirect page. And as for the guideline: using a simple slash would circumvent that problem: Diaresis / Umlaut would be acceptable within that guideline (I think). Richard 07:07, 18 September 2007 (UTC) —Preceding unsigned comment added by Richardw nl (talk • contribs)

If you're not proposing a split, then perhaps we should consider which word is more commonly used in the English language, "umlaut" or "diaeresis". Because that's the one that most readers are likely to look for. FilipeS 17:14, 19 September 2007 (UTC)

Not really the issue, since both redirect here and readers will find it. The point is that these are different things, despite being in one sense the same. If we keep them together, we don't get to choose our favourite word, we have to use both. --Doric Loon 21:30, 19 September 2007 (UTC)

Seems this subject is more sensitive than it looks ;) Look, I'm not arguing the article nor the decission to keep information on umlaut and diaeresis in the same article. That was discussed long ago. What I was trying to say is that diaeresis is the broader term. Umlaut was originally used for a specific use of a diaeresis-like diacritical mark in German. Therefore, I thought the 'official' name of the article should be diaeresis and not umlaut. Just like if there was an article 'boy' and redirect pages 'girl' and 'child' - in that case I would like to swap the function of 'boy' and 'child' as well: 'child' would be the primary name and 'boy' and 'girl' could be redirect pages. Richard 08:11, 20 September 2007 (UTC)

Except that we only have boy and girl, and no equivalent of child. You are right, though, that diaeresis is far older, going back to ancient Greek, and there is some sense in giving it priority over that young upstart umlaut. (I mean the symbol is older, not the phonological phenomenon it designates of course.) --Doric Loon 16:41, 20 September 2007 (UTC)

Umlaut marks are NOT tremata!!!

An umlaut mark which marks an umlaut and a trema which marks a diaeresis are like ß and β two completely different signs with different origins and must not be merged in one article! The same for the ellipsis (…) which is not the same as ... 91.55.87.211 17:58, 10 October 2007 (UTC)

I agree that they're not the same... but see the lengthy discussion above. I'd hate to re-open that can of worms. Richard 08:03, 11 October 2007 (UTC) —Preceding unsigned comment added by Richardw nl (talk • contribs)

Except that beta and es-zett don't have a long history of having their identities merged. They aren't completely different signs; in hot lead, they are the same piece of metal; in computer typesetting, they are the same bits in memory. I have seen no evidence that anyone has ever typeset the ö in coöperate and the ö in hör differently, excepting when one was typeset in Roman font and the other in Fraktur, which is not an umlaut-trema distinction.--Prosfilaes 15:03, 11 October 2007 (UTC)

Still, that two things look the same doesn't mean they are the same. Which doesn't mean I want to split this article up (see above) Richard 15:15, 11 October 2007 (UTC)

They don't just look the same; Unicode/ISO 10646 giving it one code means that it is the same in computers. U+0308 is the unique identifier for the character variously known as umlaut or trema.

Just as importantly, the original message's separation is wrong. The use of this mark is not merely to mark an umlaut or to mark a diaeresis. There are many languages that use the umlaut for vowels that don't have phonological umlaut, and in many cases they did, or may have, descended from the Germanic use of umlaut, not the use of the trema.--Prosfilaes 18:39, 11 October 2007 (UTC)

This may have something to do with the fact that Unicode originated from Ascii - AMERICAN standard code... - and that Americans (from the US, that is) tend to be sloppy when handling diacritical marks. A point I have made before (see section Rename Article? above) is that the name of this article is perhaps a bit ill-chosen. An argument that could be made is that in German, a diaeresis can be used to umlaut (literally: resound) some vowels and that over time the process of resounding gave its name to the diacritical mark which is used: A-umlaut which used to mean "a resounded A" now means "an A with an resound-sign" - the resound-sign being named umlaut in German. In print and/or information technology, the diacritical mark used is (almost?) always a diaeresis but in written German other marks are used as well. In other words: an umlaut might be a diaeresis but a diaeresis need not be an umlaut.

A trema is not an umlaut (which was the original statement in this discussion) since it doesn't necessarily alter the sound of a vowel BUT the diacritical mark which is used is the same: a diaeresis. So here we are: an umlaut is not a trema but they're both represented in the same way. And who knows? Maybe diaeresis is used in yet another way in some languages I don't know of. Diaeresis means much more than just umlaut or trema. That was the reason why I suggested a rename of this article in the first place.

Richard 08:18, 12 October 2007 (UTC)

Actually, the characters with diaresis in Unicode all originated from ECMA (European) standards. But good try at blaming the Americans.--Prosfilaes 14:46, 12 October 2007 (UTC)

In the beginning... ASCII didn't provide for letters like Ä, Ö, Ü so out of necessity some unusual (from a European point of view) characters ( [, |, ] ) were sacrificed. The same characters were replaced by Æ, Ø, Å in Sweden and Denmark. Likewise the British substituted # by £. These ASCII-derivates (by lack of a better term) have ISO (ISO/IEC 646) and ECMA numbers. Maybe the ECMA standarized this first, but that's most probably because the Europeans needed these characters whereas the Americans (the designers of ASCII) didn't. Since the languages that needed umlauts didn't need trema and vice versa to my knowledge no ambiguity ever arose. Therefore, like I've stated before: I don't think the fact that ASCII/Unicode uses the same mark (diaeresis) for umlaut and for trema poses a problem. I only agreed to the fact that umlaut and trema are not the same. I didn't say they weren't / couldn't be presented the same - with a diaeresis, that is. Richard 15:58, 12 October 2007 (UTC) —Preceding unsigned comment added by Richardw nl (talk • contribs) PS: when I say 'diaeresis' I mean the mark (two dots), not the phonological phenomena.

Now you are confusing things. A diaeresis is not a sign, it is the opposite to synaeresis, or literally "division" or "contraction" of vowels. Those two are distinguished by a trema which are two dots on top of a vowel. Umlaut marks have a different history; æ and œ developped to somewhat like a^e and o^e which later was simplified to a° and o° and finally became ä and ö. Tremata have always been dots on a vowel. So the fact that today umlaut marks and tremata look alike, they are not the same and don't have even a similar meaning. — looks like – or -, but those all have different meanings. Α looks like an A, but they are different letters. 91.55.106.152 19:49, 30 October 2007 (UTC)

The diaeresis is a sign. There's no question about it: http://www.m-w.com/dictionary/diaeresis defines it as a mark placed over a vowel, and Unicode uses it as the general name for the two dots. In English, it is a well established name for the physical mark.

— doesn't look like – or -; in fact, when following the second link the first thing it says is that it (the en-dash) is half the length of the first character (the em-dash). If the em-dash and en-dash looked the same, they wouldn't be distinguished. There is a long history of the use of the umlaut/trema for just about everything under the sun; any distinction by meaning, where trema means anything but umlaut, is clunky. I believe that the use of diaeresis in 20th century orthographies owes as much to German as French or any other language, whether or not they use phonological umlaut.--Prosfilaes 22:00, 30 October 2007 (UTC)

Nope, a trama is a sign to mark a diaeresis, just as a period is a sign for the end of a sentence. Diaeresis and synaeresis are different ways how consecutive vowels are spoken, and the ¨ is to mark a diaeresis so that the correct pronounciation is used. 131.220.136.195 17:15, 31 October 2007 (UTC)

You can use the words however you want, but the way they're used in English is as I said and showed. And if a trema is used to mark a diaeresis, and an umlaut sign marks umlaut, what's the name of the sign that's used in Pinyin and a bunch of other places that marks neither phonological diaeresis or umlaut?--Prosfilaes 22:41, 31 October 2007 (UTC)

In languages that use the word "trema", such as French and Portuguese, this diacritic does not always represent a diaeresis (it never does, in modern Portuguese). Examples in the article. FilipeS 02:29, 2 November 2007 (UTC)

Diaeresis usage in English

Hi,

I've added the fact tag to the In English section because they're busting my chops on this over in Heavy Metal Umlaut, and they won't let me use this article as a citation.

If someone could help me find a citation for this, I would be most grateful!

Thank you,

Webbbbbbber (talk) 17:38, 11 January 2008 (UTC)

Use in Spanish

"The diaeresis or trema is the diacritic mark ( ¨ ), used to indicate a phonological diaeresis, or, more generally, that a vowel should be pronounced apart from the letter which precedes it. That preceding letter is usually another vowel, but in Spanish it is a consonant (a g)."

Arguably, in Spanish it is used to indicate that the vowel should be pronounced apart from the following letter. Without the dieresis, the u in gue or gui is silent. With it, it is pronounced much like an English w. 66.234.222.96 (talk) 17:46, 21 March 2008 (UTC)

unnamed remarks

Hey Guys: I am a hobby Sütterlin translator from Germany, know a little Spanish and French and am quite interested in your discussion here. First things first:

"Umlaut" is a german word roughly meaning "changed sound" or "altered sound".

This is simply not true. "Umlaut" is a German word exactly meaning just the same thing as in English. No second meaning, no other connotation. To be exact, the german "Umlaut" describes the three letters ä (ae) ö (oe) and ü (ue). You can cite this from the Duden, Brockhaus or wherever you like.

Next thing is the origin of the german Umlaut, which is the e in Sütterlin, of which I am absolutely sure. In handwriting, it is still written as two vertical lines (often italic) above the letter, which looks exactly the way it was written in Sütterlin: two vertical lines (= an e) above (instead of after) a letter, which is also the reason why you can write an e after the letter instead. This letter combination (vowel + e) already existed in other languages before (e.g. Latin), but writing it like that is originally german. The two dots above a letter, however, are not. Like already said, the Umlaut is written with lines, not with dots. The dots came later, and only in printing, so today it is in most cases represented the same way as the diaeresis. --217.229.235.214 (talk) 02:31, 4 May 2008 (UTC)

1. umlaut = um + laut = changed vocal (in German one could use the word Selbstlaut). As said, it's rough but not untrue.
2. Sütterlin was developed from 1911 on, Kurrentschrift (which is also (mistakenly) called Sütterlin) is somewhat older but umlauts are even older: in the baroque times, umlauts were already used and for instance in the so-called Breitkopf-Fraktur the dots above ä are virtually indistinguishable from the one above i. So, the origin of the umlaut probably lies NOT in Sütterlin and your comment on its form is questionable.

Richard 08:48, 5 May 2008 (UTC)

Page move request (aborted)

Please see discussion here. seresin ( ¡? ) 02:00, 28 June 2008 (UTC)

use in Luxembourgish

"Since the Luxembourgish language uses the mark to show stress, it cannot be used to modify the 'u' which therefore has to be 'ue'."

what does that mean? If the Luxembourgish language uses the umlaut to show stress, an ü would be a stressed u. If you say it cannot be used in that way, it must mean that an u is never stressed in Luxembourgish. And why and when does the u have to be ue then? --androl (talk) 19:04, 16 July 2008 (UTC)

Umlaut and diaeresis

Someone please give a citation for that any use of the double-dot over vowels in non-Latin scripts is an diaeresis. This is an outrageous claim.

I usually hate people telling others to give citations, but having people telling others that the Mandarin ü is not an umlaut (lit. sound change) is getting ridiculous, especially since this is the exact same sound change as in German. This is an abuse of the word "diaeresis", and I stand by my edit that the Mandarin ü is an umlaut notation, not a diaeresis notation.—Gniw (Wing) 20:24, 5 February 2007 (UTC)

Unicode character U+03D4 is named GREEK UPSILON WITH DIAERESIS AND HOOK SYMBOL, so at least some people call it a diaeresis.--Prosfilaes 20:39, 5 February 2007 (UTC)

Yes, because in Greek it really is a diaeresis (mark indicating a separation of vowels).

The same cannot be said for Mandarin pinyin. The pinyin mark does not describe a separation of vowels, but rather a sound change (for some comprehensible meaning of "change"), even the exact sound change as the German umlaut (esp. if you consider the relationship between the Mandarin ü sounds in other dialects, or the Mandarin sounds of ü sounds in other dialects). Considering that the word "umlaut" literally means "sound change", I find it baffling that people can insist that the Mandarin "ü" is a diaeresis (which means "separation of vowels"—maybe not literally this time, but that's what it means in practice) and accuse me of "not reading the article".

Huh? I'm confused, because Greek is a non-Latin script, and pinyin isn't. If it is in fact an I-mutation in pinyin, then it should be umlaut, and should be discussed there. I don't think the article is correct in describing pinyin as a "transcription of languages that do not use the Roman alphabet"; given that it's taught in Chinese schools, it's more of an alternate script for Chinese. Please give a citation for how pinyin uses the umlaut, though.--Prosfilaes 21:55, 5 February 2007 (UTC)

In any case this also begs the question of what to call this mark in non-German Latin-script languages. To me the difference is very clear: If it describes a separation of vowels (e.g., French), it is a diaeresis, and if it describes a sound change (for any comprehensible meaning of the term "sound change", not necessarily the same kind of sound change as German—e.g., the front-back distinction of Finnish [and in fact the same front-back distinction applies to Mandarin as well]), it should be called an umlaut. Only in grey areas (e.g., the English "Brontë", which arguably could still be considered a diaeresis if we stretch the meaning to include "separation of consonant and vowel") can there be any contention. To muddy such clear distinctions is to do severe damage to the English language.—Gniw (Wing) 20:57, 5 February 2007 (UTC)

I don't think we should call it different names depending on usage. It's two dots above the letter, and I seriously doubt that many typographers actually took care to print umlaut and diaeresis seperately. Furthermore, umlaut has a fairly clear meaning--it's a specific type of sound change. It muddies that to start using umlaut to describe Malagasy language's use of it over an n to change a dental nasal to a velar nasal.

I also think that it's impossible for confusion of the names of diacritics to do severe damage to the English language. Confusing meaning changes in the word "literal", yes. Confusing meaning changes in words that most English speakers don't know because the phenomenom the word describes is irrelevant to those speakers, not so much.--Prosfilaes 21:55, 5 February 2007 (UTC)

I disagree, since the difference between umlaut and diaeresis is precisely just usage. This is why I find the Unicode Consortium's calling it "umlaut" by default so objectionable, because it causes others to think that the two are the same, when in fact they are not. To insist that we don't make the distinction simply because we assume that people are unable to make the distinction even after telling them, is not very nice.

In addition, the word “diaeresis” also has a very clear meaning. You cannot deny that by overloading the word “diaeresis” with semantics of sound change, you are also muddying things up.

With regard to whether typographers would set the two marks differently, in modern computerized typography at least, the answer would be a resounding “no”, since digital typefaces do not distinguish between the two. However, does a limitation of the tool imply some traditional distinction or usage had been invalid? Knowledge of Chinese typography, even those affecting proper punctuation, has been severely damaged by years of Western-centric computer systems; I hate to see the same happen further to English.—Gniw (Wing) 22:26, 5 February 2007 (UTC)

Unicode doesn't call it umlaut by default; it calls it diaeresis by default. The two are the same; they are two dots placed above the letter, both roughly of the same size as the dot on the i. We don't make a difference based on use any more than we distinguish the vowel y and consonant y; they're both y's. Furthermore, Unicode couldn't make a difference based on use, since Latin-1 and other previous standards didn't, and there's no evidence that users would.

Unfortunately this is not the whole story. There are many anomalies in the Unicode standard’s decisions regarding unification of glyphs. For example, Unicode distinguishes between the letter A-ring and the symbol for Angström, but the two are indistinguishable and previous standards didn’t distinguish between them, and users of Unicode are unlikely to distinguish between the two. How is this different from the umlaut-diaeresis distinction? Double standard.—Gniw (Wing) 03:11, 6 February 2007 (UTC)

You're wrong; previous standards did distinguish between the two. One of the Japanese standards had both. Even if it hadn't, a mistake is different from a double standard; you'd have to show some sort of systematic behavior to show a double standard.--Prosfilaes 14:58, 6 February 2007 (UTC)

Since both diaeresis and umlaut mean something different, we're left with the fact that there's some usage of this diacritic that is neither properly diaeresis or umlaut. To try and divide the diacritic into those two holes is thus wrong. However, the article as it is calls use of the trema to indicate a different sound, like the Albanian schwa, trema or diaeresis. (See the second paragraph.)

Why don't digital typefaces distinguish between the two? Because typographers don't care. Programs that support fine typography, like Adobe InDesign, are fully capable of letting a typographer select certain characters and choosing different glyphs, and companies producing OpenType fonts, like Adobe, can and do produce fonts that offer different glyphs. Garamond Premier Pro (pdf) offers a variety of swash glyphs and ligatures; it does not, however, offer a seperate diaeresis and umlaut.

I'm vastly skeptical that it was ever a standard practice. Traditional typography, up to the 1940s, always printed German in Fraktur. Since diaeresis doesn't appear in German and umlaut mainly appears in German, the whole word would be distinct and it would be impossible to distinguish just the two tremas.

Even if this is change, change is not all bad. Making a distinction between two virtually identical diacritics is extra work and of no use to the reader. Over the past few centuries, we've got rid of many minor things like the long s and st and ct ligatures, all of which made our writing simpler with no loss to content. There's no reason why we should still be using fonts with 5000 ligatures (like Caxton did) just because that's the way the handwritten manuscripts were when printing was invented.--Prosfilaes 00:09, 6 February 2007 (UTC)

I apologize for my mistake. But no, the distinction should still be made, because there is a distinction. We should not be so careless as the Unicode Consortium as to, for example, equate the typographically-different “ellipsis” with the “CJK three-dot leader” (we who work with mixed English-CJK typography understands; apparently other Wikipedians don’t understand, and obviously the Unicode Consortium didn’t, leaving us a mess now).

I shall substantiate my statement with an actual citation, since you all (I mean Wikipedia in general, apparently as a matter of policy now) like citations.

Firstly, from the Merriam-Webster’s Collegiate Dictionary, Eleventh Edition, 2003 (ISBN 0-87779-809-5):

• Under the entry diaeresis: [LL diaeresis, fr Gk diairesis, lit. division] a mark ¨ placed over a vowel to indicate that the vowel is pronounced in a separate syllable (as in naïve or Brontë). (p. 344; sense 1)
• Under the entry umlaut: [G. fr. um- around, transforming + Laut sound] a diacritical mark ¨ placed over a vowel to indicate a more central or front articulation. (p. 1358; sense 2).

And, for a change, from the Webster’s International English-Chinese / English-through-English Dictionary (no or lost copyright page, no or unknown ISBN, but a footnote on p. 1 indicates that it should be copyright 1954):

• Under the entry diæresis: [L. diaeresis, fr. Gr. διαίρεσιϛ, fr. διαιρεὶν to divide; διά + αὶρεὶν to take.] A mark consisting of two dots [¨], placed over the second of two adjacent vowels, to denote that they are to be pronounced as distinct letters. (p. 412; sense 2)
  [I apologize for any Greek typing mistakes; Greek is not a language I know and the words are very small in the dictionary]
• Under the entry umlaut:[G., fr. um about + Laut sound] Loosely, the two dots used in German to indicate a vowel affected by umlaut (p. 1557; sense 2)

This shows a few things:

1. My distinction between diaeresis and umlaut is well-founded
2. This distinction is at a level accessible to the general public (because it is listed and explained in an ordinary dictionary, not a special technical dictionary, book, or journal)
3. Webster’s definition of "diaeresis" has not changed substantially in 50 years, with the exception that the mark in Brontë may or may not quality as a diaeresis in the 1954 definition; however, the diacritic in the pinyin ü does not fit either of Webster’s 1954 or 2003 definitions
4. Webster’s definition of "umlaut" has changed in the intervening 50 years, with the old (1954) definition being your definition; however, according to the definition in the 2003 Webster’s, the diacritic in the pinyin ü does quite unambiguously fit the definition of "umlaut" (the diacritic), irrespective of whether the sound is a result of an i-mutation
5. The changes in meaning, at least from Webster’s’ viewpoint, follow “original intent” (the original etymology); i.e., the “diaeresis” meaning broadens to become separation of syllables (vs vowels), and the “umlaut” meaning broadens to become “front or central articulation” (vs same as a result of the German i-mutation); in other words, while the meanings of the two words have both changed, "diaeresis" still maintains the basic meaning of "division" and "umlaut" also still maintains the basic meaning of "sound change"

The 2003 Webster’s definition also unambiguously classifies the ë-as-schwa use as an "umlaut" notation, leaving only the double-dotted n in your example as unclassifiable. I personally would class this double-dotted n as an umlaut notation, broadening the "sound change" meaning further to include sound change of consonants, but obviously you won’t agree.

This would also, in a sense, substantiate my claim that dropping the proper distinction between the two marks is "severe damage", because this essentially tells people “Don’t bother to look up words in the dictionary! Even if the dictionary tells you two words mean different things, you probably can’t tell the difference so why bother?” It undermines the credibility of the precise sources that Wikipedia relies on; in a convoluted sense, ignoring the distinction is the same as destroying Wikipedia’s own sources, and so will ultimately be suicide for Wikipedia.

I just hoped that people who recklessly revert other people’s changes with just one unexplained comment should adhere to their own standards and test their own point of view.—Gniw (Wing) 02:05, 6 February 2007 (UTC)

There's nothing ambiguous about Brontë under the 1954 definition; there's no "two adjacent vowels", so it can't be diaeresis. It's really be a lot easier for me to discuss this, if you didn't rant about the off-topic three-dot-ellipsis and how a change in meaning of two obscure words can amount to "severe damage" to a language.--Prosfilaes 13:57, 6 February 2007 (UTC)

You are the one who tried to use Unicode to make a point. My comment is to show you that they are not always right and can’t be relied on as an authoritative source.

Even to prove such a little point I have to spend a few hours looking for citations and writing my reply. And what did you do? You wrote just one sentence to reject everything I wrote, not even trying to address the relevant part of my argument (which is the 2003 Webster’s definitions). You guys just rely on an uncited part of the article and reject all my comments, and pick on any small mistake I make. You guys all have double standards.

But yes, everything I say is wrong, even with my citation. Your positions are right even without any citations. I give up. I’ll just tell everyone I know not to trust Wikipedia. I quitted last year and shouldn’t have been trying to correct this in the first place.—Gniw (Wing) 14:03, 6 February 2007 (UTC)

Unicode is an international standard; the only way it gets more authoritative then that is if you can get it written into law, like the metric system has. That doesn't mean it's always right, but absolutely nothing that size is. The Merriam-Webster's 2003 isn't. (It shouldn't be called just Webster's, because anyone can use the name Webster's, and many do.) Furthermore, in the cases like the one's we've been discussion, Unicode is backward-compatible with 30 years of computing history and did it that way because it's verifiably the way it's always been done in computing.

I didn't reject everything you wrote; I wrote a short comment until I had time to write a longer one. I should have been more explicit that that was what I was doing.

But your response is what I was complaining about. When you don't immediately get your way, it's "severe damage" to the language, it will be ultimate suicide for Wikipedia, everything you say is wrong. It's not that big of an issue.

I'm not even sure this is a molehill; it's not really a factual argument, it's a definitional argument. While it's clear that your dictionaries would like to divide it in two, they leave some meanings behind, and as linked above, International Association for Coptic Studies prefers the term trema for usages in Coptic, since they don't feel diaeresis covers all their usages. People are using this term in different ways, and there is no clear right or wrong; English dictionaries describe usage, they don't prescribe correctness.--Prosfilaes 15:25, 6 February 2007 (UTC)

Hi. Gniw (Wing) and I have been discussing this in our Talk Pages. In my view, Gniw seems to be basing his objections on two ideas:

1. That the words "umlaut" and "diaeresis" primarily describe the phonetic function of the diacritic.
2. That any random fronting of a vowel can be called an "umlaut".

Both ideas are wrong:

1. The word "diaeresis" (which, remember is just short for "diaeresis mark") is widely used for this diacritic in languages where the two-dot-above diacritic does not represent a diaeresis in the phonetic sense. Two examples: Spanish and Brazilian Portuguese. This is also the case sometimes in French (e.g. aiguë).
2. The phonological umlaut does not refer to just any arbitrary vowel fronting. It's a morphological term with a very precise meaning, concerning how certain classes of words are inflected. Since Chinese is a largely analytic language, it's a safe bet that no such thing exists in it.

Let us suppose we decided to be pedantic, though, and use "umlaut marks" only for phonological umlauts, and "diaeresis marks" only for phonological diaereses. What about all the other languages that use the two-dot-above diacritic for other purposes -- including Romanized Chinese? What word would they use?... I say it's better not to confuse symbol with meaning. FilipeS 15:26, 6 February 2007 (UTC)

Let's stick to your position. What should the German diacritic be called then? It is known that the German diacritic has ceased to only indicate the "morphological" change. According to your position, even the German diacritic should not be called an umlaut.

It's not my position! I thought it was yours. My position is that, given the context, the three words are pretty much interchangeable, though in some languages there is a tradition of preferring one term to the others. I doubt that there is such a tradition in Chinese, though. FilipeS 17:15, 6 February 2007 (UTC)

No, my position has always been: it's called a diaeresis if it indicates vowel separation, and umlaut if it indicates sound change.

FWIW, like everyone that matters (i.e., every other language that use the umlaut mark to indicate front/central articulation), ü is considered a separate letter in Chinese. I stand by my position that if we have to name the two dots that make up that separate letter, it must be called an umlaut and not a diaeresis. The other position just don't make any sense at all.—Gniw (Wing) 17:06, 6 February 2007 (UTC)

Why should those languages be "the only ones that matter"?! What an arrogant claim to make! FilipeS 17:15, 6 February 2007 (UTC)

This is in the context of "it should be called a trema for all other uses". These languages don't use the diacritic to indicate syllable division, and thus the "only ones that matter" for the purpose of showing that it should not be called diaeresis for these languages.—Gniw (Wing) 17:23, 6 February 2007 (UTC)

Since this is getting nowhere, someone may go to ja.wikipedia.org and "fix" their articles. The position of the Japanese wikipedia articles for umlaut and trema (separate articles) is identical to mine (including the "the ¨ in Chinese is an umlaut" part).—Gniw (Wing) 17:31, 6 February 2007 (UTC)

If it were up to me, I would make no distinction whatsoever between "diaeresis mark", "umlaut mark", and "trema" (don't forget there are three words, not just two). A diaeresis is also a "sound change", if you're going to be that broad in your understanding of the term "umlaut", and, if you're not going to be broad, then, as I've noted several times already, most languages that use the term "umlaut" for the diacritic don't even have real umlauts.

But it was pointed out by some editors that the umlaut marks used in German and other languages related to German had historically a separate origin from the trema, and that this difference is sometimes reflected in typographical conventions, so the current division seems reasonable. I have two questions for you:

• Do you have any source to the effect that, historically, the diacritic used in the Romanization of Chinese was derived from the German diacritic?
• Do you have any source to the effect that the term "umlaut" is officially preferred to "diaeresis" or "trema" in Chinese? FilipeS 17:42, 6 February 2007 (UTC)

I will get back to you in 3 week's time, as I don't have time to do any more research in the next 2 weeks or so.—Gniw (Wing) 17:47, 6 February 2007 (UTC)

Hi folks, The article says umlaut and diaeresis marks "look similar", but I do not see any difference. I do understand the difference in function and thus in the name, but to "look similar" is to be graphically similar, while I see them graphically equal. —Preceding unsigned comment added by Gelbukh (talk • contribs) 19:49, 2 August 2008 (UTC)

Add Linux/Gnome?

Someone who have time to add in the table how to type them in Linux using the Ctrl+Alt+u+[unicode hex code]?

04:02, 10 September 2009 (UTC) —Preceding unsigned comment added by 张韡武 (talk • contribs)

Please, @ least two paragraphs, besides this phrase:

New and old forms of umlaut

hopiakuta Please do sign your communiqué .~~Thank You, DonFphrnqTaub Persina. 17:00, 8 October 2009 (UTC)

*Double Dot?

Sorry, never in my life have I heard it called this. I presume the page move is the work of some overly bored and pedantic grammar Nazi of some description.

Change it back. —Preceding unsigned comment added by 92.29.181.139 (talk) 13:54, 12 December 2009 (UTC)

Same here. I was surprised to see it moved; where was the discussion of this move? I think it should be moved back to Umlaut (diacritic) or maybe to Diaeresis (diacritic) (after noting that the Unicode standard calls uses the name diaeresis for this mark). rʨanaɢ ^talk/_contribs 21:31, 15 December 2009 (UTC)

Agreed. The terms are only diaeresis or umlaut, that must have been an initiative of just one too bold user.--Flammingo^Hey 12:36, 29 January 2010 (UTC)

Ok, now welcome to Umlaut_(diacritic), which is the subpage to the disambiguation page umlaut and linked to from diaresis, which is the phenomenon which in itself is to be distinguished from the diacritic. Any (other) thoughts?--Flammingo^Hey 13:17, 29 January 2010 (UTC) Funny, User:The Tom moved this in December without dropping a note or saying that on any of the discussion pages here, at diaeresis or diacritic.--Flammingo^Hey 13:20, 29 January 2010 (UTC)

Noüy or Nouÿ?

This is an interesting example, which I'd like to see in the French section, since it separates a vowel from a digraph. However, the placement varies between Pierre Lecomte du Noüy and Jean-Jules-Antoine Lecomte du Nouÿ; indeed, French Wikipedia says the former is a common error for the latter. Does anyone know enough about this to add it in?

Also, Ferdinand Piëch is a nice example of a true diaeresis in German, but I'm not sure on the pronunciation. [pjɛç], perhaps? (though WP-ja & WP-sr suggest [piɛç].)

— kwami (talk) 19:52, 1 October 2010 (UTC)

j + trema

It is absolutely necessary that you mention "j", precisely because it is not included in Unicode. Unicode stability policy has meant that precomposed letters+diacritics have not been encoded in almost a decade, and never will be again. As such, when you need a j + trema, you need to explicitly lose the tittle. You can't rely on a properly constructed precomposed character to take care of the situation, so this information becomes critical. It absolutely should not be stricken from this article. VanIsaac_WS 22:23, 12 August 2011 (UTC)

hi Vanisaac, All your superlatives and emphasis only make sense if you can cite actual languages, writing systems or typographical conventions where a "j" needs to receive a trema. In the absence of such reference, that mention of 'j' is pointless, and unnecessarily distracting for the reader. Best, Womtelo (talk) 12:33, 13 August 2011 (UTC).

IPA uses the the trema to indicate centralized articlulation. "j" is the palatal approximant. In dialectology or speech pathology, describing centralized articulation of the "j" approximant is indicated by j + trema. Information that is both accurate and pertinent should not be summarilly ejected from an article just because you've never heard of it. The talk page is a great place to ask these questions, but continuing to delete accurate information from an article is uncalled for. VanIsaac_WS 07:25, 14 August 2011 (UTC)

PS, if you simply did a search on none other than Wikipedia, you would see that it is used in the IPA description in the article Xhosa language. VanIsaac_WS

That's the under-trema, raised because of the descender. I've also seen [j̈] as a centralized approximant in Swedish. And even though that's composed of a dotted-jay plus trema, my IPA font knows enough to suppress the dot on the jay, so evidently this is expected enough that the font designer took account of it. — kwami (talk) 04:58, 15 September 2011 (UTC)

Requested move 2011

FYI

 – User:JCScaliger has been indef blocked as a sockpuppet of User:Pmanderson (blocked for another year for abusive sockpuppetry).

The following discussion is an archived discussion of a requested move. Please do not modify it. Subsequent comments should be made in a new section on the talk page. No further edits should be made to this section.

The result of the move request was: page moved. Consensus was clear to not use Trema. If there are still questions about the correct title, please continue the discussion and if consensus develops to change, feel free to renominate. Vegaswikian (talk) 21:28, 25 September 2011 (UTC)

---

Trema (diacritic) → Diaeresis (diacritic) – The article is currently entitled with a virtually unknown name when a much more widely known name exists. Apparently the rationale for this is that Diaeresis refers to both diacritic mark and phonetic function, whereas Trema refers to diacritic mark alone. I require a source for this claim. Every English language source that I can find treats Trema as utterly synonymous with Diaeresis; i.e. referring ambiguously to both mark and function. I believe Wikipedia is presenting a fabricated distinction between Trema and Diaeresis. It would be extremely convenient if such a distinction existed, but wishful thinking does not make it so.

In the absence of such a source, this page should be moved to the much more common name Diaeresis (diacritic).

Hesperian 10:05, 15 September 2011 (UTC)

• Oppose, though open to other suggestions. You have a point. However, this article is about the double-dot superdiacritic. That is technically two different things, diaeresis and umlaut. Therefore moving it to either one of those would be as wrong as moving Austria-Hungary to either 'Austria' or 'Hungary'.
  Other possibilities: moving it to Diaeresis / umlaut (the solution in The elements of typographic style, though generally deprecated on WP, AFAIK). Splitting the article is IMO not a good solution because half the time we have no way of deciding whether a diacritic is diaeresis or umlaut, so calling it either one would be OR.
  Please see discussions above about how if we move it anywhere it would have to be umlaut, being inappropriate to claim that some of these are diaeresis, etc. — kwami (talk) 11:50, 15 September 2011 (UTC)

Support As is clear from the very lengthy discussions above, those who have been editing this page and discussing its title are not abiding by Wikipedia's mission, which is to present information gathered primarily from secondary and tertiary sources, and not to present original ideas or research. The only issue is deciding what reliable sources say about the topic, not what they ought to say. I have taught aspects of linguistics for years, used to be on an BSI working group on transliteration, and I have never come across the term "trema", let alone used as is claimed here, and I cannot find any sources which support it (other than the current version of this article). The terms "diaeresis" and "umlaut" are used in English for both the mark and the function; this may well be considered undesirable by some editors, but this is reality which Wikipedia must reflect. So an article about the mark should be under whichever of these terms is most appropriate; in English this appears to be "diaeresis". Peter coxhead (talk) 12:01, 15 September 2011 (UTC)

• Comment The problem with either 'diaeresis' or 'umlaut' is that I don't know how you'd choose one over the other, especially when most RS's say an umlaut is not a diaeresis and vice versa. It would seem to me that 'diaeresis' violates our RS criteria. Of course, that doesn't mean 'trema' meets them. It does seem that trema is generally synonymous with diaeresis; the best I can come up with offhand is The Classical journal, 1954, p 262: "The formation of composite words and the use of the trema for umlaut are Germanic." That's not much to go on, but we have multiple RS's saying 'diaeresis' for this combined meaning is wrong.

Another, though again not much to go on: Johnson 2004:343, Bookrolls and scribes in Oxyrhynchus: "Also known as diaeresis, a trêma is in appearance a double point above a letter (like an umlaut). In payrological texts, this most often marks an iota or upsilon at the beginning of a word, though it can also (as in modern edited texts) distinguish a vowel that forms its own syllable." Now, in Greek the two words were approximately synonymous, but in English 'diaeresis' has the latter meaning. That potentially leaves 'trema' with a more general meaning. Not great, but I can't find any other word that means "two dots above a letter" without a specific function that would make it inappropriate for a title.

Consider the following, from Garry & Rubino, Facts about the world's languages : "Tone is marked in the following manners: circumflex ( ̂) is for high tone, umlaut/dieresis/trema ( ̈) for mid tone, and absence of diacritic for low tone." How can we justify putting such a mark, which functions as neither diaeresis or as umlaut, under either name? — kwami (talk) 12:11, 15 September 2011 (UTC)

Not great, but I can't find any other word that means "two dots above a letter" without a specific function. The unicode standard names these glyphs "Latin Capital letter A with diaeresis", "Latin Capital letter E with diaeresis", "Latin Capital letter I with diaeresis", etc. Considering unicode does not offer a separate set of codepoints for umlauts, those names are clear use of the word diaeresis to mean "two dots above a letter" without a specific function. Hesperian 23:08, 15 September 2011 (UTC)

We do not, in general, follow Unicode names. (For one thing, we'd have to spell them in all caps; for another, many are idiosyncratic and inconsistent with the lit.) We have numerous RS's that umlaut-a is not diaeresis-a. There are even some recommended ways of encoding the difference, though they're rarely used. What we have are two diacritics which look the same but function differently and have mutually exclusive names. There are in addition other diacritics which look the same but have yet other uses, and are neither diaeresis nor umlaut. Even if we split up the article into diaeresis and umlaut, this latter set would be left out, and we'd have the same problem naming it. Now, maybe the solution is to use "diaeresis" more broadly. Some of the people above have argued that we should use "umlaut" more broadly. I don't have a good answer, but I don't think your proposal is a good answer either.

OED, diaeresis: The sign [¨] marking such a division, or, more usually, placed over the second of two vowels which otherwise make a diphthong or single sound, to indicate that they are to be pronounced separately.

OED, umlaut: The diacritical sign (¨) placed over a vowel to indicate that such a change* has taken place.

(* A change in the sound of a vowel produced by partial assimilation to an adjacent sound)

If 'trema' is unacceptable, the only factually correct solution I can think of is Diaeresis / umlaut. — kwami (talk) 05:25, 16 September 2011 (UTC)

It seems to me your argument is circular: (a) "diaeresis" is not a word that means "'two dots above a letter' without a specific function", because (b) no reliable sources uses diaeresis in that way, because (c) any source that uses it that way is unreliable, because [goto (a)].

For the record, I was not proposing we use unicode names, and certainly not as capitalised; that is a straw man. I was merely pointing to a source that indisputably uses the term diaeresis as the name of the diacritic, regardless of function. Hesperian 06:10, 16 September 2011 (UTC)

That a rather gross misrepresentation. I never said (c). What I did say is that there is a clear distinction in the lit among diacritics as diaeresis and as umlauts, and common and repeated statements that one is not the other. — kwami (talk) 02:45, 17 September 2011 (UTC)

• Support. What is the evidence for any use of the word trema in English, let alone in this sense? The Oxford English Dictionary does not list it. Wikipedia should use a title which readers will recognize as meaning the subject.

If the choice between diaeresis and umlaut were even, we could pick one, and have a hatnote to the other. But is it? Readers who type in umlaut are likely to want the sound-shift we describe under Germanic umlaut; diaeresis can mean this mark, independently of the phonetic effect. JCScaliger (talk) 20:40, 16 September 2011 (UTC)

I don't think that's true. I've seen dozens of refs, and I assume there are hundreds, where a diaeresis diacritic "looks like" an umlaut (diacritic), and vice versa. They're not talking about sound change. Also refs that some people mistake X for an umlaut or diaeresis—again, the diacritic, not sound change. When people wish to refer to the two dots, they resort to circumlocutions like 'diaeresis or umlaut' or 'umlaut/dieresis/trema'. For a few examples of trema in English, see the ones I gave above.

We used to have this article at diaeresis (diacritic). That prompted complaints that we couldn't include umlaut, because an umlaut isn't a diaeresis. We also used to have it at umlaut (diacritic), and that prompted complaints in the opposite direction. Both names resulted in RfMs like this one because the name was "wrong". Since it's clear from the lit that these are not the same thing, I don't see how we can have both topics under one name. And since there's no reliable way to distinguish them and cover the entire topic, I don't see how we can split it under two names. Which leaves us either with a third name or a circumlocution. Now, it may be that 'trema' is either factually wrong or so obscure as to be inappropriate, but 'diaeresis' is also factually wrong, and calling an umlaut 'diaeresis' sounds quite ignorant. Certainly not what you'd want from an encyclopedia. — kwami (talk) 02:45, 17 September 2011 (UTC)

Calling the current umlaut sign a diaeresis mark is fairly common, but we don't need to go there.

The straightforward solution to that problem is to divide the article: the article Diaeresis (diacritic) can say that that its subject is a double dot, like the usual umlaut mark (and pipe to Umlaut (diacritic)). The article Umlaut (diacritic) can say that the umlaut has been marked by a variety of signs, but now usually by a pair of dots, identical to or very like the diaeresis sign, and pipe. Then both articles proceed as they are now, down to the section break; the bit on trema can be repeated if anybody can source it. So can the section on other uses for a double dot. JCScaliger (talk) 15:48, 17 September 2011 (UTC)

Where do we put all of the double-dot diacritics that are neither diaeresis nor umlaut? Keep them here? Move them to double dot (diacritic)?

Keep them in one, or both, articles. If they have English names of their own, we can make other articles for them and cross-reference; then we can use Double dot (diacritic) as a disambiguation page. "Which double dot do you mean, reader?" But if they don't (and I see little likelihood that they would be called tremata), we can bundle them into one of the two main divisions. JCScaliger (talk) 17:49, 19 September 2011 (UTC)

As for your point that calling the umlaut a diaeresis is fairly common, I agree, but then it's also common to call the diaeresis an umlaut. Seems to depend on which language the speaker is more familiar with. — kwami (talk) 01:38, 18 September 2011 (UTC)

If so, then it would be harmless to call this article umlaut (diacritic); but conflation the other way seems more common. JCScaliger (talk) 17:49, 19 September 2011 (UTC)

The above discussion is preserved as an archive of a requested move. Please do not modify it. Subsequent comments should be made in a new section on this talk page. No further edits should be made to this section.

In Swedish

I remember reading somewhere that in handwriting Swedish uses the tilde instead of the umlaut, so that ö and ä are writen by hand as õ and ã. Is this true, and if so, would it deserve a mention either here or in the article on Swedish language? --Thrissel (talk) 21:31, 29 October 2009 (UTC)

I think I have seen handwritten text by old people looking like that. Also occurring often in handwriting is using a macron ō and ā, faster to write. But handwriting can look in various ways, and this is not official. --BIL (talk) 08:44, 30 October 2009 (UTC)

As far as I know, in most languages the form of the umlaut doesn't really matter. The one exception is Hungarian since in that language ö and ü co-exist with ő and ű. Richard 09:06, 30 October 2009 (UTC)

I see. Thanks for explaining! --Thrissel (talk) 18:20, 30 October 2009 (UTC)

The Swedish "dots" can also look like an ||, //, V, U, e (turned) or ,, (I couldn't write two apostrophes). Since we don't have any letters or diacritics to confuse them with, there is a lot of room for stylistic variation for Ä and Ö. Also, with linear typefaces, using two rectangles (or lines, in thin weights), about 1/3-2/3 times higher then their width (lines in thin weights can be much higher), as "dots" is very popular in print (was even more popular before digital typesetting, since most ready made digital typefaces use different shapes, but some large Swedish newspapers and magazines have custom made fonts), but I don't think this is popular with Germans, they seem to prefer "dots" with (optically) equal width and hight, there surely isn't any widely used "international" fonts that use this type of "dots". Also, the Swedish "dots" are usually as large or larger then the dot over i and j, while Germans seem to use "dots" that usually is roughly the same size or smaller.--Se mj (talk) 22:04, 27 October 2011 (UTC)

English Ablaut

A point of clarification. The current article describes the English 'goose' ~ 'geese' alternation as perfectly parallel to German umlaut. As 'umlaut' is normally used in linguistics, however, it describes a form of vowel assimilation. This is distinct from 'ablaut', which involves changes in vowel quality to distinguish words in inflectional paradigms. (Consult the Wikipedia entries for both of these terms for elaboration.) This would make the 'goose' ~ 'geese' alternation ablaut, and in fact the example is brought up in the ablaut Wikipedia article. But would it also be umlaut? Is the idea that this is also umlaut because it represents assimilation to a now-lost plural morpheme present in an earlier form of English? Even if this is the case, 'men' ~ 'man' example would still seem puzzling, since this involves a difference primarily in vowel height rather than in backness. Could someone clear this up? If not, I'll try to remember to come back to this entry and introduce some hedging about just how similar the German and English cases actually are. MJM74 (talk) —Preceding undated comment added 19:59, 15 May 2011 (UTC).

Umlaut created the man~men contrast, and the Great Vowel Shift altered its expression. —Tamfang (talk) 02:22, 18 December 2011 (UTC)

Because Germanic umlaut is assimilation to a high front vowel (or the corresponding semivowel), it involves either a change in backness or vowel height, depending on the vowel being affected. When affecting a back vowel, umlaut is fronting; when affecting a front vowel, umlaut is raising.

If we consider a as phonologically front, the change man → men would be the raising type of umlaut, and not irregular at all. But in actual fact it may have been a combination of fronting and raising, since a was phonologically back in Old English, æ was its fronted counterpart, and e was the raised version of æ (mann /mɑnː/ → *mænn /mænː/ → menn /menː/). Or else mann had a vowel higher than [ɑ], closer to [ɔ] (since it was also spelled monn), which fronted to [ɛ]. I don't remember from my reading of Old English grammars which of these explanations is considered correct. But man → men makes sense as umlaut, either way. — Eru·tuon 08:12, 18 December 2011 (UTC)

Internal links to disambiguation pages

I just thought I'd let everyone know that the links to UMLAUT and DIAERESIS on the see also section of this article lead to disambiguation pages where the intended link was this same article. (Who managed that? xD) Someone want to change that, or should I remove them? Newwikiprofile001 (talk) 14:38, 28 February 2012 (UTC)

Replacement rules

In situations where Umlaut characters are not available, simply replacing Ä with Ae, etc. will work only for the German language. In other languages, using the underlying 'plain' character (without the "e") is more apropriate.

For example, replacing the ä's in the Finnish name Mäenpää by German rules (resulting in Maeenpaeae) renders it practically unrecognizable.--Sascha.leib 21:46, 6 July 2006 (UTC)

Agreed. The Finnish claim and example is shockingly misleading. No Finn would perform that transliteration. E.g. Kimi Räikkönen's official website is www.kimiraikkonen.com not www.kimiraeikkoenen.com (a rendering which makes me gag just looking at it) Fatphil (talk) 10:56, 9 September 2012 (UTC)

Use in Māori

In Māori a diaeresis e.g. Mäori, was used in the past instead of the macron to indicate long vowels.

This is technically true, but misleading. Taken by itself it implies that the macron replaced the diaeresis as a length marker. In fact, the macron is a long-established convention for marking length, and the use of the diaeresis for the purpose is recent – a product of an earlier phase of the computer age, when the diaeresis was was relatively easy to use on many systems, while the macron was difficult or impossible. Koro Neil (talk) 00:00, 4 April 2013 (UTC)

Proposed merge September 2013

• N - a diacritical mark is not a linguistic phenomena. Richard 08:33, 2 September 2013 (UTC)

Removed the tags. There's nothing to discuss here. — kwami (talk) 05:22, 4 September 2013 (UTC)

Printing conventions in German

Why is there an example from Finnish language under this heading? I have no expertise on this matter but I thought Finnish was unrelated to German language (not even Indo-European)157.182.105.1 (talk) 21:50, 19 June 2012 (UTC)

The Finnish and Swedish examples don't belong where they are; it's good to have a sentence or two about "omitting umlaut marks is wrong" but it distracts from a paragraph about what to do in German when the marks are not available. But where should they go? —Tamfang (talk) 23:49, 19 June 2012 (UTC)

I removed the non-German examples in the "Printing conventions in German" section. Cpastern (talk) 02:59, 8 November 2014 (UTC)

Names of the diacritic

SarahTehCat has edited the article to say that the name "umlaut" for the two-dot diacritic is incorrect. I reverted her once and asked for clarification, and she reverted me and gave an explanation in the edit summary: Because the umlaut is not a diaeresis; they mean 2 distinct phonological events. People just often get them mixed b/c it's the same character in Unicode.

@SarahTehCat: I think you're misunderstanding the purpose of the article: it's just discussing the two-dot diacritic, and it covers both the diaeresis usage of the diacritic and the umlaut usage. The specific phonetic or phonological meanings are discussed in the articles Hiatus (linguistics) and Umlaut. The reason why this article is called Diaeresis rather than Umlaut is that the diaeresis usage of the diacritic is older. Since the article is about both the diaeresis and umlaut usages of the two-dot diacritic, neither name is incorrect.

Does this explanation make sense to you, or do you still think umlaut is incorrect? — Eru·tuon 01:02, 27 May 2015 (UTC)

There are two 'identical' diacritical marks: the diaeresis (also called trema) and the umlaut. Saying that there is an umlaut in 'coöperation' is simply wrong. It might be worth considering splitting the sentence:

The diaeresis (/daɪˈɛrɨsɨs/, dy-ERR-ə-səs; plural: diaereses), also spelled diæresis or dieresis and also known as the trema, is a diacritical mark that consists of two dots ( ¨ ) placed over a letter, usually a vowel. When that letter is an i or a j, the diacritic replaces the tittle. The diacritical mark known as umlaut (/ˈʊmlaʊt/ UUM-lowt) is written in the same way.

What do you (both) think? Richard 08:24, 27 May 2015 (UTC)

Your change makes sense, but I might rephrase it to make it clear that the article is about two things: The diaeresis and umlaut are diacritics consisting of two dots.... What do you think, SarahTehCat? — Eru·tuon 06:22, 28 May 2015 (UTC)

Erutuon: That does help me understand it more, and I thank you. However, even given that nature, it is still a grossly misleading depiction of the diacritic. Because this article talks about both the diæresis and the umlaut, naming the entire article in reference to just the former implies to readers that the name "umlaut" is a synonym, which is markedly incorrect.

Richardw: I think your change is much better, but I wouldn't include the phrase "is written in the same way" because they aren't written the same way: the diæresis is typically placed over the 2^nd in a pair of successive vowels, whilst the umlaut is typically written only on the 1^st. Or are you referring to the formation of the character itself, rather than its respective functions?

– SarahTehCat (talk) 01:29, 29 May 2015 (UTC)

@SarahTehCat: So it sounds like the article title is the thing you have a problem with. One solution would be to title the article Diaeresis and umlaut or Two dot diacritic. Not sure if editors would like this solution, though. — Eru·tuon 03:46, 29 May 2015 (UTC)

@Sarah: I was indeed referring to the writing of the symbol. The diacritical mark is just that. It is not the same as the phenomena that triggers its use. If the first reference to the mark is not clear enough, consider the following:

The diaeresis (/daɪˈɛrɨsɨs/, dy-ERR-ə-səs; plural: diaereses), also spelled diæresis or dieresis and also known as the trema, is a diacritical mark that consists of two dots ( ¨ ). The diacritical mark known as umlaut (/ˈʊmlaʊt/ UUM-lowt) is written in the same way. The dots are placed over a letter, usually a vowel. When the letter involved is an i or a j, the diacritic replaces the tittle.

Richard 08:05, 29 May 2015 (UTC)

@Erutuon and Richardw: What I object to is not solely the title. Because the mark is the same regardless of how it's used, the identity—and thus the name—of the mark is defined by its function. Thus, it is only a diæresis when it is used in that function, and likewise it is only an umlaut when it is used in that function. And if, in a given situation, there is no function (say, if we're just talking about the mark itself, with no function attached)¹, then it is neither a diæresis nor an umlaut. Unfortunately, we may have to compromise in this regard since I'm not sure if there is a function-neutral name for the symbol.

__________

¹ Take the dot as an example:

1. Relative to the function of a full stop, it is either called a "full stop" (in UK English) or a "period" (in American English);
2. Relative to the function of the mathematical symbol for multiplying, it is called a "multiplication dot";
3. Relative to the function of a inter-word separator, it is called an "interpunct"; or
4. Relative to the function of an inter-numeral magnitude separator, it is called a "decimal".
5. However, when referring to merely the typographical mark itself, none of these would be accurate names for the word because the mark is not itself relative to any given function, but merely to the mark and mark alone. The mark's name would be "dot" (or some other term I'm not aware of), and only when paired with a specific function would it be called a more specific name.

– SarahTehCat (talk) 17:30, 29 May 2015 (UTC)

I'm glad you understand my point about the article's purpose. I understand what you're saying about the very distinct functions, and I'm quite aware of that. The intro paragraph should be worded so that it clearly states or implies that the article is about both uses of the two-dot diacritic, but also so that it doesn't imply that umlaut and diaeresis are the same thing. These two goals must be served at the same time. I may try to do this myself. (Also, if you don't know, it is usual to indent comments on talk pages using a colon at the beginning of a line, so that it is clear which comment you are replying to.) — Eru·tuon 02:59, 30 May 2015 (UTC)

I'm glad we can finally come to some consensus here. :) I think such a goal would be a valid one to pursue.

And what do you mean by "using a colon at the beginning of a line". If I state a person's username, isn't it clear to whom I am referring? <div style=text-align:right>– SarahTehCat</div> (talk) 17:10, 3 June 2015 (UTC)

(Okay, what the heck? My signature worked before; why's it not working this time...? —_— ) — Preceding unsigned comment added by SarahTehCat (talk • contribs) 17:13, 3 June 2015 (UTC)

ẅ

Hi everyone! I see the Letter ẅ featured prominently at the top of the page, but no further mention in the article. I have never seen this strange letter before, and I'm curious who uses it, and what for. --BjKa 08:44, 16 February 2007 (UTC)

Good question... I've deleted it, at least until someone somes up with a reference. FilipeS 14:14, 21 April 2007 (UTC)

It’s useful for disambiguating coẅorker (one you work with) from cowörker (one who orks cows). — Preceding unsigned comment added by 178.49.192.196 (talk) 06:03, 12 July 2016 (UTC)

I would suggest co-workers vs. cow-orkers ;) Richard 07:56, 12 July 2016 (UTC)

Summary: Vote

Someone please tell me WHY is the diaeresis or trema explained BEFORE the umlaut in the summary? Shouldn't the subject of the article be discussed first? Vote!

Umlaut first

• Stephen Charles Thompson (talk) 07:06, 25 November 2008 (UTC)

Umlaut second

This article is about the symbol, which in English usage is primarily a diæresis. While it is the same symbol as an umlaut, and umlaut is more common, diæreses are the (in theory) primary English usage of the diacritic, and this is the English Wikipedia. The subject is the symbol, not umlauts (which, iirc, have an article of their own that also covers sound change). Hppavilion1 (talk) 05:40, 22 November 2016 (UTC)

The article appears to equate diaresis and umlaut in the lead. Probably the subtle distinction should be clarified, and possibly umlaut not even bolded since it has its own article and is not a redirect to this article. Lithopsian (talk) 14:47, 21 February 2017 (UTC)

Was German first?

The article claim that the glyphs ä and ö have been borrowed from German into Swedish and that they started to use them from the 16th century. I have seen older texts than that in Swedish that use ä and ö with dots (denoting the same sounds as today). Usually the dots are used in small text or with Roman capitals (where rarely a superscript e is used, this is before humanist lowercase letters have been created, but the earliest humanist lowercase cursive I've seen written in Swedish (17ths century if I recall correctly), also used dots), while larger blackletter founts and writing use a superscript e. In eary Swedish printing, unlike German blackletter printing which only use Roman letterforms with words of Latin origin, you see Swedish text written in Roman capitals used for emphasis, synopsis, headers &c. In early Swedish printing, the founts are usually imported from Germany, Netherlands and Balticum, where they was created for printing in German and Dutch. The blackletter founts have professional looking superscript e's, while the Roman capitals often have clumsily added dots on ä and ö. As far as I've seen, blackletter was never used in Swedish inscriptions before the 19th century. First Runes where used and from the 15th century also Roman capitals, with dotted Ä and Ö. — Preceding unsigned comment added by 217.210.150.182 (talk) 14:20, 22 April 2017 (UTC)

Layout question

Since the infobox "Letters with umlaut" goes much further down the page than the infobox "diacritical marks|¨|caption=Diaeresis", would it be possible to put it to the right? That could look a lot neater. I tried to reverse the position of these, but had a technical problem that for some reason put one box underneath the other. Maybe someone else can work out what the problem is there? --Doric Loon (talk) 08:26, 23 July 2017 (UTC)

That list comes from {{Letters with umlaut}} which is used exclusively on this page. If I reduce the font size in that template from 20 to 10 points, it is still longer than the 'regular' infobox, but nowhere near as much. In a font size of 8 points, it is slightly shorter - but I'm not sure everybody can read it then. Richard 09:28, 24 July 2017 (UTC)

d with diaeresis

Does a d with diaeresis exist? I've seen it on a setlist of a Japanese band, there had to be a way to write it, right? Seelentau (talk) 17:10, 2 November 2017 (UTC)

Since one can use a diaeresis to compose characters, one could place it just above anything. That does not necessarily mean it exists in any real language. Furthermore, the two dots are used to represent other phenomena as well. In this case, it could be a metal umlaut. Richard 08:00, 3 November 2017 (UTC)

Yes, but what I mean is, this character needs to exist in some way for them to be able to type it on a computer, right? Seelentau (talk) 02:35, 5 November 2017 (UTC)

There is no specific Unicode character for it (although there is a combining diaeresis character), but you should really be asking general questions at something like WP:Reference Desk/Language or possibly at Computing if it's more that aspect. This talk page is for discussion about improving this particular article. --Deacon Vorbis (talk) 02:44, 5 November 2017 (UTC)

@Seelentau: not necessarily. Using 'composing characters', you can put those dots above anything. You could even put it above numbers or characters like '@', '#' and '-'.

@User:Deacon Vorbis: the question is directly pertinent to the subject of this article. I don't see the problem in asking this question here. Richard 08:50, 6 November 2017 (UTC)

Chloë and Zoë: borrowed names

" Chloë and Zoë" are written like this not only because of the silent "e" but also because they are directly borrowed from French (with Greek roots) and it kept the French original form. Same thing for "Noël" and "naïve", they are directly borrowed from the French, and kept their genuine spelling. The article is not complete if it doesn't mention that the trema in those words is caused by borrowed words. There fact that the "e" would be silent is not the cause without it, but it could be the cause for the conservation of the spelling. Please, complete the explanation. — Preceding unsigned comment added by 78.250.168.130 (talk) 05:20, 13 December 2017 (UTC)

The use of both, however, is considered to be largely archaic.

If it doesn't mean "of both at the same time", a better wording would be "of either" --Backinstadiums (talk) 18:41, 30 September 2019 (UTC)

Hiatus

Spanish uses it in poetry to indicate a hiatus pronunciation of a diphthong or triphthong --Backinstadiums (talk) 18:44, 30 September 2019 (UTC)

Omissions?

Is there some reason D, F, and G have been omitted from the chart? (D̈,d̈,F̈,f̈,G̈,g̈)

24.136.4.218 (talk) 12:59, 14 March 2021 (UTC)

Probably because they don't exist as precomposed characters and that no language uses them. Using the same technique as you did, I can put two dots over a grinning cat emoji but that doesn't make it meaningful. --John Maynard Friedman (talk) 14:26, 14 March 2021 (UTC)

Can also be below the letter z̤

Unicode calls z̤ which I found on Romanization_of_Persian#Main_romanization_schemes LATIN SMALL LETTER Z and then COMBINING DIAERESIS BELOW.

This article doesn't even mention that diaeresis can be below a letter. Akeosnhaoe (talk) 03:01, 31 January 2020 (UTC)

@John Maynard Friedman: you just reverted an edit that added COMBINING DIAERESIS BELOW to the \{\{diacritical marks\}\} template on the article. Did the Unicode consortium call it the wrong thing? Akeosnhaoe (talk) 18:33, 20 August 2021 (UTC)

This character is listed on Combining Diacritical Marks and it links to this article. Should that link be removed as well? What is this thing's correct name then? Akeosnhaoe (talk) 18:40, 20 August 2021 (UTC)

Ouch! As it stands, the article says They both consist of two dots ¨ placed over a letter, which is why I reverted. I have no idea what two dots underneath is called.

@Nardog and LiliCharlie:, can you help? (or know someone who might?) --John Maynard Friedman (talk) 19:16, 20 August 2021 (UTC)

Sorry, Akeosnhaoe, I failed to read your comment properly, specifically the reference to the Unicode code-point COMBINING DIAERESIS BELOW. So if the IP-editor's change is reinstated, then the opening sentence needs changing to match? How? --John Maynard Friedman (talk) 19:26, 20 August 2021 (UTC)

The IPA symbol name is "supersubscript umlaut," see Handbook of the International Phonetic Association (1999:181). — Always ping me if you want my comment; I no longer watch the English Wikipedia. Love —LiliCharlie (talk) 19:41, 20 August 2021 (UTC)

@John Maynard Friedman: I don't know what it should be changed to. On the one hand it seems to me that it would be more accurate to say that a diaeresis/umlaut is the two dots regardless of whether they were placed over or under a letter, even though in practice they're effectively always placed over. On the other hand, the only example of "real" usage that I gave is by a committee inventing a romanization scheme for Persian in 1999 and also, if I write a cedilla above a letter for some reason, you would just say that I wrote a "cedilla over a letter", even though cedillas go under letters. Akeosnhaoe (talk) 20:19, 20 August 2021 (UTC)

Yes, that is fair. So how does this read?

The diaeresis (/daɪˈɛrɪsɪs/ dy-ERR-ih-sis (also known as the trema or tréma) and the umlaut (/ˈʊmlaʊt/) are two different homoglyphic diacritical marks. They both consist of two dots ¨ placed over a letter, usually a vowel. When that letter is an i or a j, the diacritic replaces the tittle: ï.
A similar diacritic, placed underneath the letter, is used in specialized contexts and is not typically encountered elsewhere.

in the lead; and in the body:

==Diacritic underneath==
The International Phonetic Alphabet specifies a 'subscript umlaut' [example needed]; and the ALA-LC romanization system specifies it as a variant pronunciation of Z, and is used, for example, in one of the Romanization_of_Persian#Main_romanization_schemes. In Unicode, the code-point for this diacritic is U+0324 ̤ COMBINING DIAERESIS BELOW.

Comments? --John Maynard Friedman (talk) 20:55, 20 August 2021 (UTC)

See Breathy voice#Transcription for IPA examples. In IPA superscript and subscript umlauts are used contrastively; the former means "centralized", and the latter "breathy voiced". A breathy-voiced central open vowel transcribed as [ä̤] makes perfect sense, and it wouldn't be surprising to encounter this sound in some speakers' accent, perhaps not even in English. — French Wikipedia article fr:Ṳ mentions several orthographies where ⟨u⟩ with subscript umlaut is used: « Ṳ, ou U tréma souscrit, est un graphème et une lettre additionnelle de l’alphabet latin utilisée dans l’écriture du kayah, du minbei, du mindong et du puxian, ainsi que dans l’alphabet scientifique des langues du Gabon (ASG) de 1989. Elle est formée de la lettre U diacritée d’un tréma souscrit. » Love —LiliCharlie (talk) 21:26, 20 August 2021 (UTC)

@LiliCharlie: Merci! Yes, I see that notation documented in en.wiki too, at Red Karen language (aka Kayah) for example. I will look at the others tomorrow. It might mean my draft will need further development. --John Maynard Friedman (talk) 23:26, 20 August 2021 (UTC)

P.S.: The examples in the list on page 181 of the IPA's Handbook are ⟨b̤ a̤⟩ (as in the IPA chart), and the example on page 25 ("Exemplification of the symbols" section) is: "Hindi [kʊm̤ar] ‘potter’". Love —LiliCharlie (talk) 02:29, 21 August 2021 (UTC)

I have added some text to the article (three places) to describe double dot underneath. Doubtless it can be improved as I have no linguistics expertise, so feel free. --John Maynard Friedman (talk) 11:35, 21 August 2021 (UTC)

References

Are diaeresis and umlaut necessarily homoglyphic?

On German requirements Crissov reports that certain fonts provide "localized glyphs shapes" for German. Also I imagine that in a handwritten German phrase like schöner Citroën some use a "-shaped umlaut for ⟨ö⟩ but two dots for the ⟨ë⟩ diaeresis. Does someone have sources for such variation? And if not, do we have a reliable source that says the diaeresis and the umlaut are invariably represented by identical glyphs? Love —LiliCharlie (talk) 16:00, 21 August 2021 (UTC)

Oh, I just found this with illustrations of localized glyphs for Ä, Ö, Ü. Love —LiliCharlie (talk) 16:04, 21 August 2021 (UTC)

On Every character has a story #24: U+0308 (COMBINING DIAERESIS) Michael S. Kaplan cites the German Institute for Standardisation as follows: "Currently, a substantial amount of existing German data distinguishes between Umlaut and Trema. Both diacritics have a similar, but not necessarily identical representation..." That doesn't look like we can fully maintain our "homoglyphic diacritical marks" claim in the introductory sentence. Love —LiliCharlie (talk) 16:55, 21 August 2021 (UTC)

Although the Deutches Institut für Normung asserted that "both diacritics have a similar, but not necessarily identical representation", they don't say (or at least Kaplan doesn't report them as saying) how they differ. DIN 5007 is not freely available but multiple sources say (for the purposes of sortation), the user may (a) just ignore the diacritic or (b) resolve it to a suffixed e (and ß to ss) – so the case can't have been an easy one to sustain, which is how we ended up with the ISO/IEC recommendation in our article (which I have just cited using the Kaplan paper you found, thank you).

Homoglyphic (as you know) means 'same glyph': the glyphs are the same. The problem described in the DIN submission seems to be a semantic one: depending on language the diacritic has a different meaning. I have to agree with ISO/IEC that this is not a typographic issue. If it walks like a duck and quacks like a duck, then it is a duck.

I am reminded of the fuss when the Consortium declared many Chinese and Kanji characters to be identical, that the apparent differences were merely stylistic, and thus they should use the same code-point. They were not happy but failed to produce convincing evidence to the contrary.

It is true that we don't have an RS that says that they are homoglyphs but neither do we have one that says that they are not. I don't think we need a citation in this case per wp:the sky is blue.

There is some material on German handwriting at https://german.stackexchange.com/questions/32628/can-an-umlaut-be-written-as-a-line-in-handwriting but I don't see that it supports any distinction (even if it were an RS, which it isn't). --John Maynard Friedman (talk)

I forgot to comment on the Github article on "German requirements": the article is clearly about typeface design and the mechanics of creating a computer font. Whether a type designer chooses to render U+00C4 Ä LATIN CAPITAL LETTER A WITH DIAERESIS as something like ˙A˙ or ˚A˚ rather than the conventional Ä, that is surely not something we should take into account, any more than we should È being routinely printed as E' in Italian or Ó as O' for Irish names – or even the widespread habit of discarding diacritics completely!

--John Maynard Friedman (talk) 21:44, 21 August 2021 (UTC)

Right, I neither meant a difference in encoding, nor one in collation, but a glyphic difference in the same text and the same typeface (which is unlike Han unification).

Dutch Wikipedia article nl:Trema mentions the possibility of such a difference: "In handschrift en in een zorgvuldig ontworpen lettertype kan er een verschil te zien zijn tussen een deelteken en een umlaut. De umlaut wordt soms als twee korte verticale streepjes genoteerd. Typografen verkiezen soms de puntjes of streepjes van een umlaut wat dichter boven de letter te plaatsen dan de puntjes van een trema." Note that the umlaut and the trema can occur in one and the same German or Dutch phrase without any problems, as in the German schöner Citroën or the Dutch Schröder in België. I don't think the two look the same is as obvious as the sky is blue, simply because in German and Dutch that doesn't seem to be the case, neither in traditional/careful typography, nor in all handwritten styles. Love —LiliCharlie (talk) 23:40, 21 August 2021 (UTC)

P.S.: The Github article on "German requirements" is about adding umlaut glyphs to a font that already contains trema glyphs, and access them via OpenType features. Love —LiliCharlie (talk) 23:51, 21 August 2021 (UTC)

Printed text from 1763 with German in one font and Latin, French and digits in another.

In modern writing, diæresis and umlaut are usually homoglyphic; however in some German fonts and scripts used IIRC until WW2 the umlaut could be represented by two parallel vertical bars, and in some older fonts by a small raised e. These never happened for the diæresis. German doesn't use diæreses, and until the first half or so of the XX^th century German text (possibly including umlauts) was printed in a Fraktur or similar font while French, English, Spanish, etc., quoted words (possibly including diæreses), or translations from German into these "non-German" languages, were printed in an Antiqua or similar font, even on the same page, see Antiqua–Fraktur dispute; this is a traditon several centuries old, see illustration at right (with ^e for umlauts).

So nowadays diæresis and umlaut are usually homoglyphic; but there is no necessity for them to be so. — Tonymec (talk) 23:53, 21 August 2021 (UTC)

1. Diaereses are used in German, e.g. in proper names such as Piëch and Hoëcker (which avoids mispronunciations due to the sound values of the digraphs ⟨ie⟩ /iː/ and ⟨oe⟩≈⟨ö⟩ /œ/). 2. Should we then rewrite the introductory sentence to "... are two different but usually homoglyphic diacritical marks"? Love —LiliCharlie (talk) 01:30, 22 August 2021 (UTC)

I've just remembered reading Sütterlin and sure enough, the umlaut was written as " not ¨:

 

So the sky is not blue! I now think we should not say 'homoglyphic' at all. And an explanatory section in the body is definitely required. --John Maynard Friedman (talk) 11:25, 22 August 2021 (UTC)

I had always assumed that diæresis marks in German would only be used in loanwords. And conversely for umlaut marks in French, so that there would be no natural co-occurrence. So the counter-examples of nl België and de Hoëcker have left me mind-blown. Now I wonder, how historically recent is the use of two-dots-above to indicate diæresis in languages that also have umlaut? ⁓ Pelagic ( messages ) 14:05, 22 August 2021 (UTC)

@Pelagic: It blew me away too: I've had quite a few 'everybody knows' preconceptions blown away in the past few days! I'm sure that a more informed editor can give the long answer but from my reading, it is the two-dot umlaut that is late to the party, not the other way round. Historically, the name Müller (as in Müller rice would have been written something like Mu^eller and got 'rationalised'. --John Maynard Friedman (talk) 15:52, 22 August 2021 (UTC)

Evolution of umlaut?

I've been reading DE:umlaut and came across this evolution of ⟨ä⟩ from Latin script to modern German type. No superscript e (though the body of the article supports the 'e on top' version). Curiouser and curiouser. --John Maynard Friedman (talk) 19:24, 22 August 2021 (UTC)

The illustration to the right does show superscript e. The e following a in line 2 and the one above a in line 3 take the same shape, except that the full baseline form also has "combining"/"linking" left and right strokes. Love —LiliCharlie (talk) 21:44, 22 August 2021 (UTC)

Encoding of trema vs. umlaut

@John Maynard Friedman: Unicode explains how to distinguish the two here. The proposal they adopted is this one. Note that the unified character is referred to as combining diaeresis in the proposal (and the rejected character as combining umlaut), whereas the functionally distinguished diacritics are referred to as tréma and umlaut. Love —LiliCharlie (talk) 02:50, 23 August 2021 (UTC)

Draft rewrite of opening sentence

Given the 'difficult' history of this article, I thought it best to propose here first a draft rewrite of the opening sentence in the light of the the discussion in this section. As of today, the lead sentence is:

The diaeresis (/daɪˈɛrəsɪs, -ˈɪər-/ dy-ERR-ə-sis, -⁠EER-) (also known as the trema or tréma) and the umlaut (/ˈʊmlaʊt/) are two different homoglyphic diacritical marks. They both consist of two dots ¨ placed over a letter, usually a vowel. When that letter is an i or a j, the diacritic replaces the tittle: ï. The same mark, placed above or below the letter, is used in other contexts and for different purposes and meanings.

How do fellow editors feel about this replacement:

The diaeresis (/daɪˈɛrəsɪs, -ˈɪər-/ dy-ERR-ə-sis, -⁠EER-) (also known as the trema) and the umlaut (/ˈʊmlaʊt/) are two different diacritical marks that (in modern usage) look alike. They both consist of two dots ¨ placed over a letter, usually a vowel. When that letter is an i or a j, the diacritic replaces the tittle: ï. In computer systems, both forms have the same code point (binary code). Their appearance in print or on screen may vary between typefaces but not normally within the same typeface. The same mark, placed above or below the letter, is used in other contexts and for different purposes and meanings.

My draft drops the (french) tréma, replaced 'homoglyphic' with 'that (in modern usage) look alike' and added "In computer systems, both forms have the same code point (binary code). Their appearance in print or on screen may vary between typefaces but not within the same typeface. ". Comments? --John Maynard Friedman (talk) 13:23, 22 August 2021 (UTC)

(random drop-in) I like it! ⁓ Pelagic ( messages ) 14:09, 22 August 2021 (UTC)

Since then, I have changed it slightly to add a 'normally' (in "not normally within") and added a footnote to recognise that nothing is truly impossible. --John Maynard Friedman (talk) 15:41, 22 August 2021 (UTC)

I shall have to change the footnote because one may easily recreate the Sütterlin style using 'combining double vertical line above'. Later. --John Maynard Friedman (talk) 18:24, 22 August 2021 (UTC)

The choice of the correct glyph is not a matter of encoding, and Unicode's private use area (PUA) is an outdated way to handle glyphic variation. In modern fonts, several glyphs can represent the same encoded (Unicode) character. The choice is then determined by language markup or other ways to trigger the OpenType features the font designer has chosen (as illustrated here using CSS, but you can also do that in LibreOffice and other word processors). — In Wiki projects, language markup is automatic, so en.WP has markup for English, de.WP for German, etc. You can override default language markup using CSS or templates like {{lang}} (use the italic parameter if necessary). Of course users need a suitable font for the desired result, but if they don't they still see something that's reasonable, which is not the case if PUA characters are used. Love —LiliCharlie (talk) 22:10, 22 August 2021 (UTC)

Yes, I agree completely. My quick "I need to revise the footnote" comment was because I had realised exactly the same thing. It gets worse: I had strayed into OR and NOTGUIDE territory and definitely need to discard that footnote. So the revised version reads simply:

The diaeresis (/daɪˈɛrəsɪs, -ˈɪər-/ dy-ERR-ə-sis, -⁠EER-) (also known as the trema) and the umlaut (/ˈʊmlaʊt/) are two different diacritical marks that (in modern usage) look alike. They both consist of two dots ¨ placed over a letter, usually a vowel. When that letter is an i or a j, the diacritic replaces the tittle: ï. In computer systems, both forms have the same code point (binary code). Their appearance in print or on screen may vary between typefaces but not normally within the same typeface. The same mark, placed above or below the letter, is used in other contexts and for different purposes and meanings.

Better? --John Maynard Friedman (talk) 22:43, 22 August 2021 (UTC)

Perhaps I see things that aren't there, but font designers who follow some typographic tradition don't do anything that's not normally done in their time-honoured culture, so to me, it seems less ambiguous and more neutral to replace "not normally" with "rarely" or something similar. The rest sounds fine. Love —LiliCharlie (talk) 00:42, 23 August 2021 (UTC)

Austrian teaching script

I just found this graphic that shows "Austria's standardized cursive" according to German orthography#Alphabet, and the "Österreichische Schulschrift seit 1995" if de:Schreibschrift#Widersprüche im Verständnis von Schreibschrift can be trusted. Anyway, it seems to be a contemporary script used for teaching purposes in Austria in which the tittles of i and j are true dots but the umlauts are lines. That makes me wonder if it's not misleading to say that the diaeresis and the umlaut "(in modern usage) look alike" and that "both consist of two dots". Maybe we should find a wording that reflects that the umlaut can still be written as two strokes, and that schoolchildren in parts of the German-speaking world are still taught to write that way. — However this is the point where things get really confusing, because in Hungarian (Hungary and Austria used to form a single country) there are two umlaut marks: Two dots are used for short vowels, and two strokes for long ones... Love —LiliCharlie (talk) 06:23, 23 August 2021 (UTC

@LiliCharlie: Yes, that is helpful. The article on de.wiki refers to Austrian practice but (depending of Google Translate) I struggled to make sense of it. It also mentioned the Hungarian use of 'double acute' diacritic. I'll change 'normally' to 'rarely', specify 'in type' and put it live later today. Writing the body text is going to be fun! --John Maynard Friedman (talk) 09:07, 23 August 2021 (UTC)

Our article Double acute accent#Handwriting also says that "[i]n handwriting in German and Swedish, the umlaut is sometimes written similarly to a double acute." Unicodewise, double acute remains the wrong choice for German and Swedish, of course, no matter how the umlaut is represented graphically. Love —LiliCharlie (talk) 10:07, 23 August 2021 (UTC)

Double dot, umlaut, diaeresis, etc... Time to split this article?

I was originally in favor of keeping Umlaut (diacritic) and Diaeresis (diacritic) a single article, distinct from the article Diaeresis concerning the term from prosody/phonetics, but this seems to be swimming against the tide. Users keep adding remarks about the diacritic to the Diaeresis article, which I've given up on removing. It would be nice if we could use the word Trema for the diacritic (Greco-Latin + Germanic), Diaeresis for the prosodic/phonetic term and Umlaut for the grammatical/linguistic/Germanic term. Unfortunately, the word "trema" is rare in English, and we must use the most common terms. So here's a different idea:

• Diaeresis (diacritic) → Diaeresis; let the article describe both the phonetic change and the diacritic which often represents it;
• Umlaut (diacritic) → Umlaut; let the article describe both the Germanic grammatical feature and the diacritic(s) which has historically represented it; include also the languages whose spelling was influenced by that of German.

Not that by splitting Diaeresis (diacritic) and Umlaut (diacritic) into different articles we won't be saying that they're necessarily different diacritics (which is an unimportant academic question, at any rate). But we will be acknowledging that both terms are used in English, sometimes with different (historical or unhistorical) connotations.

A problem remains: where to place things like Pinyin, where the diacritic has no historical connection whatsoever with neither a diaeresis nor a grammatical umlaut, or languages like Albanian, for which it's unclear whether their spelling was most influenced by German or Greco-Roman precedents? This will be a little tricky, but it could be solved with some common sense and some flexibility (allowing some overlap between the parts of the articles on Umlaut and Diaeresis that deal with spelling, if necessary). FilipeS (talk) 13:22, 20 June 2010 (UTC)

And what of the double-dot in Russian? I'm not sure how we'd work all those out, or how we could expect the reader to know which was where.

Why not trema? : "a diacritic consisting of two dots ( ¨ ) placed over a letter, used among other things to indicate umlaut or diaeresis". That would avoid the more specific connotations of the other two words (diaeresis in poetry, umlaut in sound change), even if it's not as common (though I'm not sure it's much less common that "diaeresis"). Etymologically, the word refers to the pips on a die, and so could more easily be either.

I've also gutted diaeresis into a dab. There was almost nothing there but the diacritic.

We don't have to use the most common term. We commonly avoid it when it would be unduly ambiguous or misleading. ("ambiguous or inaccurate names for the article subject, as determined by reliable sources, are often avoided even though they may be more common. For example, tsunami is preferred over the arguably more common, but less accurate tidal wave.") This is precisely the kind of situation where a less common word may be preferred.

Well, since there haven't been any other comments in 3 mos, I'll go ahead and move. Just revert me if I'm being overly bold. — kwami (talk) 19:52, 1 October 2010 (UTC)

---

I can't believe that umlaut (diacritic) redirects to this page. No wonder people are confused! I'm going to find out how to get the two pages demerged. Dadge (talk) 16:37, 14 August 2020 (UTC)

I support this as well. Let me know if there's anything I can do to help! Webbbbbbber (talk) 22:38, 17 August 2020 (UTC)

It is ridiculous that most of the time spent discussing this article is spent on how to differentiate between the diaeresis and the umlaut, a problem that would largely be solved by having separate articles. It appears that a few people are being intransigent, and some of those people don't even know German! Dadge (talk) 10:55, 7 October 2021 (UTC)

See WP:splitting but especially WP:PROSPLIT. I assume you don't plan to use the 'be bold' option, so you will need to open a new discussion topic at the bottom of this talk page. I strongly recommend that you draft the proposal in your sandbox, so that you have fully worked out which bits go in which article and which in both – and why – before uploading it here. Good luck! --John Maynard Friedman (talk) 12:48, 7 October 2021 (UTC)

Opening summary

After describing the two uses of the diacritic, the summary has a paragraph that reads, "Neither of these phenomena occur in English, except in loanwords (like naïve) or for stylistic reasons (as in the Brontë family or Mötley Crüe)." Yet the subsection "diaresis" under "Names" directly contradicts that statement, as does the discussion of English under "Modern Usage".

I recommend the sentence of the opening summary be rewritten: "In English, the use of the mark for Umlaut exists only in loanwords such as Fahrvergnügen. As diaeresis, its usage in English is largely obsolete apart from loan words (such as naïve) and to clarify certain proper names like Zoë. The mark is also used simply for style, such as "Mötley Crüe")."

Also, the use of "Brontë" is an example of using the mark to make sure the e is pronounced, which is a (unusual) form of diaeresis in that it separates a consonant and a vowel, not two vowels.

-- G001 (talk) 17:12, 7 March 2022 (UTC)

Has anyone else ever heard of wikt:Fahrvergnügen being used in English? Can anyone suggest anything better? (I don't think that Müller Rice counts.) Wikitionary says that it was used by Volkswagen in advertising, which rules it out to my eyes.

the use of "Brontë" is an example of using the mark to make sure the e is pronounced is reported frequently, but doesn't make any sense to me. 'Brone' or 'Brote' might be rhyme with 'loan' but 'Bronte' could only be BRONtay or BRONty. It was an affectation created by Bronte Père, a social climber who thought the Irish name 'Brunty' would hinder his advancement. Linguistically, it never made sense, it seems to me to be no different in reality to the metal umlaut (like "Mötley Crüe").

I can't see where the existing text contradicts the #Names section?

I can't see a contradiction with #Modern usage either (I assume you mean the para beginning "In Modern English, the diaeresis, the grave accent and the acute accent are the only diacritics used apart from loanwords."), except to the extent that the sentence It may be used optionally for words that do not have a morphological break at the diaeresis point is not reflected in the lead but I would argue that this is a rare and optional exception to the general rule. But maybe we should include it anyway.

The current sentence is no more than the latest attempt to give a succinct distillation of some complex material, particularly for ESL readers – see #Confusing and poor definition above – so any improvement is welcome provided that it is better. If we can find a better example than Fahrvergnügen then yes, do it. --John Maynard Friedman (talk) 17:47, 7 March 2022 (UTC)